Identifying a target market is the first and most important step in building a marketing campaign. For every campaign created, a target audience is required. Machine learning enables marketers to discover the right targeting criteria automatically without needing to manually select targeting criteria. In this post, we run through how marketers use machine learning to automatically create campaigns and the targeting criteria associated with those campaigns.
Building Target Segments Without Machine Learning
Let’s step through how a marketer would typically build a target segment of users. As an example, a retailer might be interested in running a promotional campaign for all users who are interested in shoes. But how does the retailer know who is interested in shoes? It might be based on whether they purchased shoes in the past. It might be based on whether a user has visited a shoes page on their website. The retailer is left with a subjective decision around which criteria to use in creating a target segment of users.
Building Target Segments With Machine Learning
Enter machine learning. With machine learning, the criteria used to build a target segment are automatically selected based on the goal. If the goal is to find users who are interested in buying shoes in the next week, we can set that goal and machine learning will automatically identify the features most important to that goal. All available customer attributes are used, which allows the marketer to understand the most important attributes across all user attributes.
In Cortex, Feature Importance ranks your customer attributes by importance and determines which features provide more value than others. Some features are more important, but all features provide some level of value to our predictions. Using machine learning for campaign targeting allows you to step up your targeted marketing strategies.
Building your own ML pipeline to predict the most important targeting attributes of your customers is simple. Watch the quick demo of Cortex below.
Curious in learning more about building predictive Machine Learning Pipelines in Cortex? Email us at email@example.com or fill out our contact form!